1. AMPA Receptor Tetramerization Is Mediated by Q/R Editing
Ingo H Greger, Latika Khatri, Xiangpeng Kong, Edward B Ziff 
Neuron 
Volume 40, Issue 4, Pages (November 2003)
DOI: /S (03)

2. Figure 1 The Reentrant P Loop Mediates ER Retention
(A) Left: schematic of the GluR2 subunit. The P loop is shown in red, the Q/R editing site as a blue diamond, and the lipid bilayer as dashed lines. Transmembrane segments are labeled, TM1, 3, and 4. Right: sequence alignment of the P loop, and adjacent C-terminal residues, between GluR2 and the KcsA K+ channel. Residues mutated in this study are marked with asterisks. Conserved residues are shaded yellow, conserved residues are shaded green, and Arg607 at the Q/R site is shown in blue (triple asterisks indicate that multiple mutations were introduced at this site).
(B–D) Pulse-chase analysis of GluR2 pore mutants. Hippocampal neurons (∼10 div) expressing Myc-tagged GluR2 mutants were pulsed with [35S] Met-Cys for 20 min and chased for 5 hr (see Experimental Procedures). Mutants were immunoprecipitated with anti-Myc Ab, EndoH digested, and analyzed on 6% SDS-PAGE followed by flurography. (B) Mutants in lanes 3 and 4 were made in the Q607-background. (C and D) Mutants at position 607 are labeled in single letter code; in (D), position 611 was additionally altered to Asn (lanes 6 and 8). Maturely glycosylated GluR2 is denoted with a filled arrowhead, immaturely glycosylated GluR2 with an empty arrowhead.
(E) Cell surface expression of Myc-tagged GluR2 pore mutants. Hippocampal neurons (20 div) were stained live with monoclonal anti Myc-Ab (to detect surface Myc) followed by fixation, permeabilization, and staining with polyclonal Myc-Ab (to detect total Myc). Myc surface expression was measured as the ratio of surface to total cell area. Bar graph shows mean ± SD; n = 30. Significance was determined with paired Student's t tests (two-tailed distribution) *p < ; **p < Mutants are indicated below the x axis.
Neuron  , DOI: ( /S (03) )

3. Figure 2 Molecular Model of the GluR Pore Region
The model is based on atomic coordinates of the KcsA K+ channel, and spans region G580-M650 of GluR2. Left top panel: reentrant P loop (residues F592–S616); the ascending pore helix is in red, the descending selectivity filter region in gold; side chains discussed in the text are shown in stick representation. Right top panel: two adjacent subunit fragments are shown in red and blue, in side view. Long helices correspond to TM3, which are analogous to the inner helices in KcsA. Q607 is shown in stick. Bottom right panel: side view of the tetramer; two adjacent subunits are shown in red and blue. Bottom left panel: top view of tetramer shown on the right. Q607 residues, shown in stick, are near the center of the ion conduction path.
Neuron  , DOI: ( /S (03) )

4. Figure 3 The Q/R Site Determines AMPAR Assembly
(A and B) Sedimentation analysis of GluR2(R) and GluR2(Q). Hippocampal neurons expressing GluR2(R) and GluR2(Q), respectively, were [35S] Met-Cys pulse labeled for 20 min and chased for various times (shown on the left). Triton X-100 cell lysates were separated on 10%–50% glycerol gradients; gradient fractions were subjected to anti-Myc IP and analyzed on 6% SDS-PAGE and flurography. The direction of sedimentation is indicated by an arrow. P1 and P2 denote the two major GluR peaks.
(C) Quantification of the kinetic analysis shown in (A) and (B). Signals were quantified using a PhosphoImager (BioRad). GluR2-Q signals (time points 1, 5, and 13 hr) from fractions 4–16 were summed up and set at 100% (total); individual signals in each fraction were expressed as percent of the total.
(D) Steady-state sedimentation analysis of various GluR2 pore mutants (indicated on the right). Signals were scanned densitometrically and quantified using NIH-Image software (quantification was as described in Figure 3C).
Neuron  , DOI: ( /S (03) )

5. Figure 4 Analysis of Assembly States within P1 and P2
(A) EndoH analysis of P1 and P2 gradient fractions. GluR-Q (chased for 5 hr) was immunoprecipitated from P1, and P2 peak fractions with Myc-Ab, and EndoH digested. GluR2-Q in P1 is mainly EndoH sensitive, but EndoH resistant in P2.
(B) [35S] Sulfate-labeled GluR2-Q predominantly sediments in P2. Neurons expressing GluR2-Q were [35S] Sulfate labeled for 13 hr. Triton X-100 cell lysates were fractionated on a 10%–50% glycerol gradient. Signals were quantified as described in Figure 3C.
(C) AMPAR assembly occurs in the ER; blocking ER exit with Brefeldin A does not impede assembly into P2. (Top) BFA inhibits maturation of GluR2-Q (lane 2). After infection with GluR2-Q, neurons were incubated with 1 μM Brefeldin A (BFA) for 13 hr. BFA was also present during Met-Cys depletion, and the 5 hr chase. Lane 1 (−BFA), lane 2 (+BFA). Samples were analyzed as described in Figure 1B. (Bottom) GluR2-Q-infected (BFA treated) neurons were lysed in CHAPS buffer, and lysates analyzed on 10%–50% glycerol gradients (as described). BFA treatment was as described above (C). Fractions 4–15 were immunoprecipitated with anti-Myc Ab, flurographed, and quantified as described in Figure 3C; mean ± SD; n = 2.
(D) DSP crosslinking of peak fractions. P1 and P2 fractions (indicated on top) from a 10%–50% glycerol gradient, were crosslinked with 2 mM DSP for 30 min on ice. Crosslinked lysates were separated on 6% SDS-PAGE and analyzed by anti-Myc Western blotting. Size markers (in kD) are shown on the right. Black arrowheads mark positions of mono- (M), di- (D), and tetramers (T).
(E) BN-PAGE of peak fractions. 30 μl of glycerol gradient fractions (indicated on the top) were taken up in BN-loading buffer, separated on a 4%–12% Bis-Tris gradient gel, and analyzed by anti-Myc Western blotting. M, D, and T on the left side denote monomer, dimer, and tetramer, respectively. As a size control, F#11Q was treated with 1% SDS for 10 min at rt to dissociate the tetramer; the resulting product was designated as monomer (n = 1). The graph shows how gels were calibrated with native protein size markers, to estimate sizes of the dimer and tetramer relative to the monomer (see Lee et al., 2002). Markers, denoted as white squares, were—thyroglobulin, Mr = 669,000; ferritin, Mr = 440,000; catalase, Mr = 232,000; lactate dehydrogenase, Mr = 140,000; BSA (bovine serum albumine), Mr = 66,000.
Neuron  , DOI: ( /S (03) )

6. Figure 5 Endogenous AMPARs Sediment in P1 and P2
(A) Sedimentation analysis of endogenous GluRs from cultured neurons. CHAPS lysates were separated on 10%–50% glycerol gradients. Gradient fractions were analyzed by Western blotting using Abs indicated on the right. GluR1 and GluR2 distributions within the gradient were determined by densitometric scanning of autoradiograms using NIH image software (bottom panel). Quantification was as in Figure 3C (mean ± SD; n = 3).
(B) EndoH treatment of fractions from Figure 3A. Fractions were immunoprecipitated with anti-Myc Ab, EndoH treated, and detected with anti GluR2 Ab. GluR2 in P1 is immature, but mature in P2.
(C) BN-PAGE analysis of AMPAR assembly states in P1 and P2. Gradient fractions from Figure 5A (indicated on the top of the gel) were separated on a 4%–12% Bis-Tris gel as described in Figure 4E. Samples were loaded in duplicate, and analyzed by GluR1 (left) and GluR2 (right) Western blotting. GluR2 monomers are abundant in P1, the dimer/monomer ratios (D:M) are indicated on the bottom of the gel.
(D) Assembly kinetics of [35S] Met-Cys labeled endogenous GluR1 and GluR2, chased for 5 hr. CHAPS lysates were separated on 10%–50% glycerol gradients. Subunits were immunoprecipitated from gradient fractions; GluR1 top, GluR2 bottom, and analyzed by 6% SDS-PAGE and flurography. Solid arrows mark the immunoprecipitated subunit, empty arrows mark coprecipitating subunits. Quantification (bottom panel) was as in Figure 3C (mean ± SD; n = 3). The experiment was repeated in the presence of BFA to block mature glycosylation (bottom gel). Shown is a GluR2 IP.
(E) GluR subunit-subunit associations predominantly exist in P2. GluR2 was immunoprecipitated from fractions across a 10%–50% glycerol gradient. GluR2 associations with GluR1 (top) and GluR3 (bottom) were assessed by Western blotting. The bottom panel shows a longer exposure; note the wider spread of GluR3 in P1, extending into F#6.
(F) GluR1 and GluR2 sedimentation properties also differ in adult rat brain. A 350 × gav PNS was separated on a 10%–50% glycerol gradient, and subunit distributions analyzed by Western blotting; GluR1 top, GluR2 bottom. Note the spread of GluR2 into F#5 and 6, which is not seen for GluR1.
Neuron  , DOI: ( /S (03) )

7. Figure 6
Arg at the Pore Loop Apex Attenuates Tetramerization of GluR1 and of Heterologously Expressed GluR2(R)
(A) Neurons expressing HA-GluR1 (wt Q600 and Q600R) were [35S] Met-Cys pulse labeled and chased for 5 hr; CHAPS lysates were separated on a 10%–50% glycerol gradient and analyzed by 6% SDS-PAGE and flurography as described in Figure 3 (mean ± SD; n = 3).
(B) Myc-GluR3 assembles into P2. Neurons expressing wt Myc-GluR3(Q) were [35S] Met-Cys pulse labeled and chased for 5 hr; CHAPS lysates were separated on a 10%–50% glycerol gradient and analyzed by 6% SDS-PAGE and flurography.
(C) Arg607 determines assembly in HeLa cells. CHAPS lysates from HeLa cells expressing GluR2-R and -Q were separated on 10%–40% glycerol gradients (this range was chosen for increased resolution of P1). Fractions were separated on 6% SDS-PAGE and analyzed by anti-Myc Western blotting. Note that glycosylation-induced shifts of GluR2 are more pronounced in HeLa cells, and mature receptor in P2 is evident without prior EndoH treatment (bottom panel; F#12, 13).
(D) Suggested rules for the assembly of AMPAR dimers. Left panel (Di) shows disfavored subunit arrangements, such as GluR2 homomers (left), and possibly tetramers containing three GluR2 subunits (right). Middle panel (Dii) shows favored tetramers, such as receptors containing a single GluR2 subunit (left), and diagonally positioned GluR2 (right). Right panel (Diii) shows a tetramer assembled from a GluR2 homodimer. R stands for GluR2-R, Q for GluR1, 3, or 4, and the blue diamond depicts Arg607.
Neuron  , DOI: ( /S (03) )